Slashdot Mirror

Implementation is total FUD. I agree totally that the things said are totally off the wall. However, we are dealing with a criminal justice system that a) upheld the one-click patent, b) does not care one lick about how silly implementations get, c) (to paraphrase someone whose name escapes me) decides who has the better attorneys and not necessarily who has the more correct position, d) has rules written by people who brought you the DMCA, e) does not care about the original intent of the internet.

I, myself, think that the ideas I presented are worthless from a technical prospective. However, neither you nor I are judges. Most judges do not care how a problem is solved as long as it is solved. Fact is, this is may become a reality.

Until all expert witnesses that can be called agree with you--that there is absolutely no way under the sun to limit transmissions on the internet--and this is proven in a court of law (not in a journal), we should worry.

At most, half is FUD. The fact you believe all of it is FUD illustrates the main point I want to make: according to precedant in the US, the right to free expression is subordinate to the right to commerce on any given medium.

Wake up and realize that what I am saying is ridiculuous only because you don't believe it can happen. It can happen. The government can come in and regulate anything they so choose for as little a reason as a damn good lawyer can defend. That is the paradigm you should be thinking in, not that of protocols and workability.

PerES Encryption

Your adversarial tone implies to me that you do not understand the mechanism by which this can be done. Let me explain...

Let's have a router at some IP with a netmask of N that routes traffic for some entity(s) [Say from a network of subscribers to @Home, my provider]. Simply drop all outgoing SYN traffic not destined for IP addresses not registered with the federal government. Don't tell me it can't be done. I do it at home with a 486 running Linux. I just do it backwards. When done backwards, it's called a firewall. Routers don't care where packets originate from (hence end-2-end), they just decide where to send them to get to their destination. There is no reason why they cannot be filtered by an access list.

PerES Encryption

People who claim that Gnutella and Freenet can't be shutdown are kidding themselves. The way they will be shutdown will be more horrific than any kind of censorware out there. There will be outrage and there will be shock at the step and those who think that this scenario won't happen are kidding themselves.

Follow my logic. The legal system is all about suing who has the money. Even those who have a passive role in the commission of a crime are obliged to pay damages. Who has money and facilitates the commission of these crimes? ISPs. With the advent of "technologies" like Carniwhore or POS-2000 or whatever its name is, those who are in court will realize (mistakenly) that ISPs can filter information passing through them. Thereby, injunctions will be slapped on the big ISPs like @Home, etc. (but not AOL, for some mysterious reason). The ISPs will then start filtering for known patterns of bytes of Freenet, Hotline, etc. traffic and block them in either direction. Of course, this solution is ridiculuous to think of, but then again, judgements of law are often unencumbered by the thought process.

Of course, there are obvious ways around this. They will be implemented until the ultimate work-around (use of encrypted packets) at which point entire ranges of ports will be banned. Probably, even worse--everything but port 80 from a list of "registered web servers".

If you think this is absurb, try this on for size. Broadcast something for 24 hours at about 100MHz. Yep, that's right, the FCC will be on you in a heartbeat to shut you down.

To think that these cannot be shutdown is absurd. To think that the government will not try to regulate the Internet in an absurd fashion is hubris. They have done it before (from the sinking of the Titanic onward, the US has regulated airwaves) and they WILL do it again.

The fact that we have licensed radio stations is proof enough for me.

PerES Encryption

try this instead:

<# echo bash > output
># echo "#include " > output.c
># echo "int main ( int argc, char **argv ) { system ("bash"); }" > output.c
># gcc output.c -o output

I think you'll find you get my point.

PerES Encryption

Is it a bug in Gnutellish clients that data gets transferred? Seriously.. The fact that there is no signature on any Gnutella packet decrying the type of client being used, how could this be fixed? As far as the gnutella spec reads, at no point does it rely on a human to directly respond to each query. Rather the queries are assumed to be xmitted to clients. This just happens to be a client that is not intentionally run.

It conforms (mostly, it seems) to the spec for xferring data. That makes it a valid gnutella client. Without a montioring of the type of client sharing data, there is no fix.

In other words, this is as much a bug as typing:

$ su - root
# echo bash > output
# chmod 777 output
# chmod u+s output

and expecting the operating system from preventing attacks. It is not a software nor an implementation problem. Rather, it is an attack on the protocol that relies on human engineering to work. (ie, Gnutella operates on a big fundamental flaw.. all clients are kind and good)

The way I see it, it was just a matter of time. Those who wish to transfer data anonymously should consider the source of the data. Fact is, unless you can authenticate the source, then expect garbage and get surprised from time to time.

In other words.. I double dog dare somebody to fix this in software. And even if they manage by some stroke of super-genius to fix it, it will not prevent similar attacks entirely.

PerES Encryption

Many aspects of linguistics overlap CS.. particularly the facility with which concepts can be expressed--for this I cite Noam Chomsky (sp?).

Other aspects of information translation figure into CS--ie NLP, the holy grail of user interfaces. A good reason why this should fall under CS is that CS provides a vehicle (ie Turing machine and others) for which it is possible to prove certain things as correct in an abstract way--ie without many preconceptions. All that must be taken for granted are the existence of zero(0), the existence of a next number, a way to express whether or not a number is larger than another, and 3 truth tables (and, or, and not). From these six things, all concepts that can be explained to a computer can be explained in terms of these 6 axioms--albeit, it gets complicated.

Of course, the problem lies in the location of actual information--idiomatic, ambiguous, and connotive meanings are the tricky beasts. For instance if I say "I cleaved the fat from the meat," it has a completely different meaning from "I cleaved the fat to the meat." The meaning of the word "cleave" and the information the meaning carries does not reside with cleave--since cleave is its own antonym. Rather, it lies in the prepositional phrase modifying the verb. Of course it can get more complicated (Hofstadter has quite a few fun ones in his books).

Computers, however, do not deal with ambiguity very well. They need 100% perfect transfer of information as well as intent to operate within constraints specified.

Without knowing a priori what is acceptable to a civilization and what is not (it was not uncome for mothers to fellate their sons as a sign of affection in fuedal China, but a kiss on the cheek was considered intimate), information and intent must be passed 100% correctly. It seems to me that CS is a natural candidate to study this.

PerES Encryption

Excluding multi-process/multi-thread timing conditions, politics of design, promises made by salespeople, and ill-formed specifications, this is a reason why systems crash.

My current project at work has a very fluid specification. I cannot tell you what is garbage and what is good for a particular module until the module is written and given for inspection to those who are designing the overall process to use the software. That means I have to make guesses about something I know nothing about so as to make a design that will work with something. All I have are guidelines. Should the guildelines change a week before delivery, in goes the fix--sans testing and sans rational design philosophy. In a word, I knowingly write shoddy code to fit political design decisions to meet deadlines set by ill-advised individuals who know not the first thing about how a computer operates.

Quite possibly, the single biggest problem with coding is the concept of a deadline. The second, almost because of the deadline, is the "I'll figure that out later" philosophy. Without a full blueprint of an application or system, (hell, I'd be happy with an accurate spec once in a while) it is impossible to make sound decisions. Ergo, bad software.

What you speak of are luxuries to your average code-for-a-living individual.

PerES Encryption

An inventor too poor in capital to exploit his invention directly is free to approach potential partners under terms of non-disclosure. This is protection of the inventor's interest via a private contract, not a patent.
I tell you how to make a widget after signing an NDA, you tell a third party. You are liable, not the third party who makes them. NDAs are good but rely on people keeping secrets. You and I both know that people cannot keep secrets. If I wanted relief, I could only sue you. You probably don't have fortunes to recover potential losses. Can't get blood from a turnip. Probably end up costing me more to sue than what I would get in the end.

If the idea truly requires much development and assembling of resources to exploit then the originator will have the natural advantage of a significant head start and perhaps deserves no more than that.
Most of the problem with production is not the production itself, but securing capital and real property to do the construction. Production is (generally) very quick since Henry Ford. What takes so long is building a place to work, securing the necessary assecories (machines, computers, etc.) to work, and staffing. That is generally a large barrier of entry. After this is the cost of raw materials and salaries. Once again, I need to protect myself from companies that already have infrastructure.

If the idea is so easy to copy that others can very quickly bring a competitive idea to market after the inventor's version becomes public then maybe it was not such a brilliant idea after all and does not deserve the benefit of a legal monopoly.
So, would you have one company collecting these ideas so that you purchase everything from safety pins (patented last century) to computers? The theoretical (not as it stands today) point of a patent is to protect innovation. The fact that something is "simple" is not mentioned--only non-obvious. Most magic tricks are simple to perform but are intentionally non-obvious. Bad example, but you get the point.

The basic question which needs to be addressed, not begged, is: How do we know when the natural commercial advantages that accrue to an originator, however large or small, are insufficient and must therefore be supplemented by a legal device such as a patent?
This is the most eloquent statement of the connundrum I have ever read on Slashdot. Due to the quick nature of these comments, I am often too heavy handed. By the nature of putting the word "legal" in the sentence, the subjective words "advantages" and "sufficient" need to be precisely defined for what I see as an advantage and sufficient, would differ from your opinion. This seems to me to be a restatement of the halting problem. Is it possible to predict the future or the impact of innovation?

As for Thomas Jefferson, he was approaching the issue of IP from the standpoint of moral philosophy, not economics -- and even if he was agruing economics what does his personal finances have to do with it? Would you have us believe that Bill Gates is a better economist that Milton Friedman, Paul Samuelson and all the other Nobel Prize winning economists put together because he is wealthier than they?
Without getting too far away from the topic, Friedman, Samuelson, et al. did not die homeless. Rather, they had a keener understanding of personal property. And I would much rather have Herr Gates run a company I owned then any other person in the past fifty years except maybe Sam Walton. Go back 100 years, Gates is still my pick over Rockefeller. Go back 1000 years and I think the original Hapsburg might have him beat.

On the theory of business--go to a businessman. On the theory of economics--go to an economist. On the thoery of political (cough) science--go to, well, a philosopher. A philosopher is only as good as his assumptions.

PerES Encryption

To quote TJ as a good source of info on economics is like quoting Mel Brooks for history. He died penniless. At one point he tried to raffle off his house to pay his debts. Couldn't raise enough money. He also had serious problems with the national budget (before we had a standing army and before the SSA, and before NASA) This guy had his head in his proverbial arse when it came to understanding business and economies. Rather than assail his credibility directly, I offer this:

and it may be observed that the nations which refuse monopolies of invention, are as fruitful as England in new and useful devices

Of the countries in the 18th century that are as "fruitful" as England was, I wonder why it was England who managed to hold on to an empire (a measure of success in the 18th century) that spanned the world until as recently as 1997?

I am wondering right now, what country on the planet is as "fruitful" as the United States? Let's take for example, China. There is no such thing as intellectual property in China. Yet they managed to gain about 30 yrs in their missle program in the past 10. Is this due to the fact that their freedom of intellectual property (as long as that property is deemed fit by the party), or that MIRV technology was given to them so they could launch satellites?

How about the Soviet Union of the 60s, 70s, and 80s? While it is true that they could maintain with the US up to a point, has the general populace of Russia benefitted from the old CCCP space program? Contrast with America for homework.

Think about it in terms of "brain drain"--the demographic principle that states intelligent people migrate from places where there are few opportunities. They migrate to places where there is opportunity. What is this opportunity? The opportunity to succeed--defined however you like. Most people define this as financial security--especially those from poorer portions of the world.

Onto the heart of the argument:

If nature has made any one thing less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of every one, and the receiver cannot dispossess himself of it

Suppose you have an idea. Just that, an idea. Let's say that your idea is to build a wonder widget. To build this, you would need real property (in the accounting sense) to implement your idea. Further, and most often the case, you would need some help. These two things can be very expensive. Then there is R&D and other intangible expenses. You go to the SBA and get a loan. You pitch the widget, they love it, and loan you $10M seed.

You proceed to build/rent/lease/buy the real property necessary to produce your widgets. You hire technically trained artisans to build your widgets. You get some bright lights to run the R&D, accountants, managers, etc. Without producing one single widget, you have spent a good bit of your $10M. You purchase raw materials.

Acme, Inc. comes along and manages to get a hold of some plans for widgets. With their infrastructure, they gain quickly and soon surpass your production of widgets. Their crack team of salespeople start distributing widgets globally. You, on the otherhand, have no distribution model, have little capital, are in debt, and will probably fail, go bankrupt, and pay debts for the rest of your life.

The sole purpose of a patent is to present a barrier to entry into a market. You do not own your idea, but rather have an exclusive lease. Admittedly, the time given for the lease is not as optimal as it could be. However, most people who start companies need this barrier to get moving. Without this barrier, it would be a guarantee that there would be only one company in the world.

It is true that the "receiver cannot disposses himself of [the idea]," but the patent system is there to prevent that receiver, be it Joe Schmoe or Ultra-Mega Corp, from profitting from the use until you have had a shot at profitting at it. You, as a relatively poor entity, cannot compete with a super corporation. That supercorporation could just take out your livelihood.

I believe that these are valid points and stand against your everyday biggoted, incestuous pervert (not to impeach your source's credibility) who was an exceedingly poor business man and died in debt (that part was to impeach his credibility).

The lesson you should walk away with is that patents prevent monopolies in an overwhelming majority of cases. For every bad patent, there are 100 good. This isn't to say that there are no bad patents, but let's not throw the baby out with the bath water.

PerES Encryption

LaTeX still does not solve three basic problems with document formatting. After working with JNI to implement printing in Java (Java printing is absolutely horrid), actual font sizes vary from printer to printer. Courier 10 is different sizes on different printers--can you imagine the headache? Actual points on a page vary from printer to printer (albeit slightly) across platforms. Believe it or not, a dual boot 95/NT machine will render a document differently on the same printer with the JNI class we implemented--depending on the OS being used. The final problem is cross-platform implementation of these issues.

To suggest LaTeX or other page layout that is not as complete as postscript or pdf, is to trade the language, not the problem.

The popularity of HTML is that it is easy to learn the basics and rich enough to do very complicated formatting. Eventual HTML standards may approach TeX 3.14159 with things like MathML, Style-sheets, etc. However, there will always be vendor specific problems. For those who look at my web page, I made it in 30 minutes with vi. I do have a very good working knowledge of HTML.

XML may be a way to go. As an XML programmer, I enjoy the fact that I have lex and yacc built in with XML, though I don't like the bloat. The project UIML makes the point I am trying to make. The interface should be easy to use such that any graphic designer can use it and those of use who code and hate doing UI can go about coding something real. Keep it simple yet rich. Avoid scripting et al.

Anyway, back to work on what my web page describes.

PerES Encryption

The problem is this.. what is legit? If it were so easy to say s/he is legit, there would be no reason to have, say, a judicial system. The judicial system is there to judge the credibility and context of evidence. If, for instance, you and a pal enjoyed playing unreal. After taking a serious whipping, he emailed you to say "Ha! Ha! You suck!" and you replied "I'll kill you tomoroow!", things can get taken out of context. If your pal shows up with a kitchen knife in his throat the next day and Carnivore was sniffing your email, you may become suspect numero uno.

So you're innocent, but you just happen to be out taking a spin by yourself during the time of the murder. Hard to defend alibi. You could conceivably become indicted. That costs money. Whether or not you are innocent, it costs lot's of money (lost wages, bail, lawyers, etc.) and the information still has to be judged in your favor for you to be cleared since it is pretty obvious you made a death threat as revenge for some strife. If the pocketbook argument doesn't work, think about your personal credibility. If that doesn't work, remember bail is not usually given in murder one cases in my parts. Perhaps a suspension of your civil liberties may convince you.

You must remember that it is the context not the content that determines legitimacy. Carnivore can capture tons of content. However, it is impossible to ensure that it captures enough content to discern context. In some cases, like the one I mentioned above, it is impossible for it to determine any context. The English language (as a matter of fact, all sufficiently complex languages) is open for interpretation and your interpretation of a harmless note is not always the easiest to believe.

Was it Ben Franklin who said something to the effect: "Those who would give up liberty for safety deserve neither liberty nor safety."? The fourth and fifth amendments are not there to harbor criminals. They are there to protect the wrongly accused. And just because you claim legitimacy does not mean that you will never be accused. Go ahead and let them sniff? Given enough time and wide enough scope, the FBI could have brought charges against Mother Teresa.

Rather than a court order allow a switch to be thrown, I would prefer a larger price in time and money to install such a system to deter wanton use of this if it can even do what it claims. Remember, it's not how private something really is in the US, but how much privacy you expect that determines what kind of warrant is needed.

PerES Encryption

Is consciousness computable? Is it recursively enumerable? Is it even algorithmic?

Are people speaking of something they know little or nothing about?

A Turing machine is used to determine the answers to these questions--not to model anything. TMs came before computers (which, by the way, are most accurately modelled by LBAs for which there is a solution to the halting problem).

A Turing machine is best used to gauge the theoretical efficency of algorithm as well as give a solid framework to what an algorithm is. The limitations on the TM are not those of the computer but rather the mathematical (ie totally theoretical) limitations and carry the weight of expressions like 1 + 3 = 4. Due to the way these symbols work, it is not valid for 1 + 3 = 5. There is no real reason, only mathematical axioms that hold. Same with a Turing machine. Things proven with a Turing machine hold theoretically.

Computers are not TMs but rather an approximation of a certain small group of TMs (particularly the TM which represents the universal LBA (linear bounded automaton)).

To make broad sweeping statements that intelligence is or is not recursive, recursively-enumerable, or undecidable is a bit premature since we cannot accurately describe the problem nor can we accurately describe the solution in terms that are acceptible for use with the mathematical concept of a TM. My current belief is that it is undecidable (ie, non-algorithmic). This does not mean it cannot be duplicated by man in a lab, just that the TM model of computation cannot represent it. Follow my reasoning:

1. Intelligence is not an algorithm to enumerate a set of correct solutions. In other words, intelligence does not have a final answer, but rather an evolving set of current "good enough" conditions from which to operate.
2. There is no accepted "yard stick" for intelligence. Without a way to accurately measure intelligence, without an accepted standard definition of intelligence, and without a method to test solutions to problems for intelligence (as opposed to luck or misapplication of faulty intelligence), there can be no way to mathematically determine whether or not intelligence exists.
3. There can be no accepted measure of intelligence. I define intelligence here as using intuitive (non-algorithmic?) processes to exercise better judgement. "Better" is subjective and philosophers have been arguing for thousands of years whose judgement is better. As a matter of fact, if a yardstick for intelligence could be developed, it would finally finish what Godel started 80 years ago in that philosophy as a study will be as useful as astrology. Simply take the conclusions of two philosophers, measure the intelligence, and take the more intelligent. Eventually, it would evolve into a more concrete science like astronomy.
4. The exercise of intelligence often comes with experience. Therefore, there is no agreed upon initial state, since those excersing intelligence have diverse experiences. Therefore, by definition, there is no single state from which intelligence arises. These experiences could be in the womb or genetic factors inherited or any number of things. The fact is, there is no "good" starting point--once again a subjective that cannot be measured.
5. A TM requires (among lots of other things) two very special states: initial and halt. By the points raised above, these states cannot exist.
6. Therefore, a TM cannot be constructed to manufacture intelligence.

There are those who would say that birth and death are pretty good initial and halting states. However, no two people are born the same and death is an artificial consequence of being alive not being intelligent (trees also die).

I am not saying that intelligence cannot be duplicated by man. I am just saying that current models of computation cannot do it. Just because a car can move you from point A to point B does not mean that point A and point B can be on two different planets. The mechanics that make up a car cannot accomplish this just as the current models of algorithms cannot model intelligence. A radical change in thinking is required. Whether or not that change will come is still in doubt.

PerES Encryption